[R-sig-Geo] problems with predict() in raster package

Häring, Tim (LWF) Tim.Haering at lwf.bayern.de
Thu Feb 3 10:31:50 CET 2011


Hello,

I get an error message when using the predict function (raster package) with a GLM and a rasterStack. 
Did somebody have the same error message or can give me a hint what I am doing wrong.


> dv <- raster("dv.sgrd")
> fpi <- raster("fpi.sgrd")
> hut <- raster("hut.sgrd")

> train <- read.dbf("train.dbf")
> str(train)
'data.frame':	2574 obs. of  4 variables:
 $ id : int  26 32 51 71 77 82 97 110 112 118 ...
 $ dv: num  0.507 0.454 0.51 0.422 0.465 ...
 $ fpi: num  0.968 1.396 0.974 1.343 1.412 ...
 $ hut: num  0.465 0.723 0.662 0.4 0.742 ...
 - attr(*, "data_types")= chr  "N" "N" "N" "N"

> myglm <- glm(dv ~ fpi + hut, train, family=gaussian())

> r10 <- stack(fpi, hut)
> r10
class       : RasterStack 
nlayers     : 2 
nrow        : 9736 
ncol        : 7762 
ncell       : 75570832 
projection  : NA 
min value   : NA NA 
max value   : NA NA 
extent      : 4494280, 4571900, 5446480, 5543840  (xmin, xmax, ymin, ymax)
resolution  : 10, 10  (x, y)

> p <- predict(object = r10, model = myglm, filename='pr10', progress='', na.rm = FALSE, overwrite = TRUE)
Error in endrow - nrow : non-numeric argument to binary operator


Thank you very much
TIM

--------------------------------------------- 
Tim Häring
Bavarian State Institute of Forest Research 
Department of Soil and Climate
Hans-Carl-von-Carlowitz-Platz 1
D-85354 Freising

E-Mail: tim.haering at lwf.bayern.de
http://www.lwf.bayern.de



More information about the R-sig-Geo mailing list